So Markdown is this Lightweight Markup Language. Everyone (relative; among programmers, writers, and other “power-users”) uses it. LLMs use it. So it’s destined to eat the world. But it doesn’t mean Markdown is good.
↫ Artyom Bologov
We have these crazy fast and complex computers, but I’m still supposed to style text with obscure, arbitrary symbols, like an animal? We invented WYSIWYG decades ago, and our computers should be able to figure out how to properly share styled/unstyled text without us users having to learn markup languages using arcane symbols that require weird claw grips to type.
The widespread use of Markdown is not indicative of its merits; it merely underlines the utter failure of the computing industry to fix basic problems.
I am in love with what markdown plus pandoc have given me.
Mind you, I mostly use plain text based tool because of what I do (programming). I favor everything that I can read directly on my editor of choice. So for documents that do not need anything fancy, it is great. If I just need simple stuff, like paragraphs, several levels of titles and the odd table, I can write that in markdown focusing on content, not form, using the same tool I use for my other work: a text editor and integrating wonderfully with my workflow, where git and diffs happen as routine.
The documents can later be converted to a pdf or something else to please clients, and in that step I can put some extra time to improve the form of the already completed content. I also like the separation between the source document (the markdown) and the target document (usually a pdf).
I could be using latex or something like that, but markdown is enough most of the time and ir is very legible in its source form.
I’d even say that WYSIWYG has resulted in so many flaws. So many documents that may look fine but are internally flawed in its structure. Formatting via white space and newlines that break with different tools…
WYSIWYG has resulted in many being able to use the tools badly. They are good in the sense that they provide a nice learning curve, the problem is that many just don’t take the steps needed to do things properly (like using proper styles instead of ad hoc formatting).
IMO, for many use cases the text based tools are much supperior as they focus more on content and less in form. An for that tools like markdown (or rst) provide a very low barrier of entry option.
You somewhat summed it up why we never get a reliable working WYSIWYG editor. Because why the hassle as a dev when you anyway write code in syntax.
I very much disagree. The whole notion of using a tool like word, with it’s myriad ribbons and buttons, is now an anachronism. You ( or you and an AI or just the AI ) will create content. There maybe some light styling to indicate headings and lists and such in the content – as markdown offers – but otherwise there is very little need for any additional styling, especially because you have no notion of the medium used to consume the content.
The consumers AI will then reformat your content, and very likely edit it for them based on their preferred way of consuming that content. The consumer may not even read it at all but simply have the AI tell them about it or make a tiktok like video about it.
The future of content is you talking to an AI and it putting it together, and for the purists perhaps a totally distraction free typing experience where it’s just you and your thoughts on a blank page. It’s certainly not hundreds of buttons just in case you want an italicized underlined heading at 24pt and the user consuming that on an HTML page with ads everywhere. That world is dying, and good riddance .
I would love to have both for my documentation. Nice syntax behind a reliable WYSIWYG interface.
Word (processors) is (are) at least as arcane as markup language.
Sometimes I just want to focus, run my computer without any graphical interface, and can work on some code or write and format some document without any visual distractions.
I agree that GUIs could be way better, but the arcane art of using a computer like an animal has its uses.
Markdown is my favorite thing invented in the recent past. I never want to go back to using shitty WYSIWYG editors or heavy markups.
Markdown is magical because it can be rendered, but looks great as plain text, which is how I mostly interact with it.
It is used for doc comments in modern programming languages and it easily beats HTML in this role, which was often a historic choice.
Thom Holwerda,
This almost sounds like an argument in favor of using AI to stylize text without users needing to manually format the text. IMHO this seems doable.
Osnews has decades worth of articles, have you ever wondered what would happen if you used that data to train an AI? Not trying to suggest there is a good commercial reason to do it, but I’m often intrigued by these kinds of ideas.
Edit: Hopefully asking these questions doesn’t come off as provocative, but wouldn’t you be curious to see what the AI can learn from your own works?
I don’t need a computer to regurgitate stale shit at me when I am perfectly capable of producing new shit myself.
Thom Holwerda,
Sure, I know your take on LLMs to generate content. However you expressed a frustration “our computers should be able to figure out how to properly share styled/unstyled text without us users having to learn markup languages”, which is a problem that AI might be able to solve by analyzing content you’ve previously created to help you apply styles to new content.
I understand the more widespread public rational against AI that’s trained on public data and/or takes away jobs, but is it fair to say that your philosophical opposition to AI extends beyond this, even if AI is using your own data to do your own job? The question that comes to my mind is this: who is this unfair against?
“our computers should be able to figure out how to properly share styled/unstyled text without us users having to learn markup languages using arcane symbols that require weird claw grips to type.”
Sorry, we are way to busy dazzling you with shitty pseudo-AI that does not work as expected to bother making anything useful. -Software companies
This article shows a fundamental misunderstanding of *why* Markdown exists and was happily welcomed.
Non-technical writers welcomed a way to write and save text documents in a manner that 1) works with any tool that accepts plain text, and 2) does not lock you in to a tool. Markdown is not intended to replace LaTex, or HTML, or Word .DOCX. It is the equivalent of .RTF with the *important* benefit of being human-readable.
Why does this matter? Because sometimes you don’t have full control over every device you want to write on. Many of us have day jobs with locked-down desktops. We work across platforms/OS’s. We like using vintage computers alongside modern computers. We like writing in “distraction free” or simplified toolsets. Our needs are fairly simple, but yes, sometimes we really want to be able to indicate **bold** or *italicized* words (which seems to horrify the author of the above article.) So, without Markdown, what are our options?
– RTF fits the bill is many ways, and is helped by being an older format. This means you can use word processors dating back to the 1980s for many platforms (DOS, Mac, Amiga, etc). But — it excludes pure text editors (including sending documents in plain text emails. It is also NOT human-readable.
Every other non-human-readable format suffers the same problem, made worse by either proprietary lock-in or by the fact that older editors and word processors won’t be gaining support for new formats, ever.
– HTML would be preferred by the technical cognoscenti like the article’s author – but it is far harder to learn than Markdown, and far messier to read. The core Markdown syntax does not rely on “code” – it simply unifies a set of very readable plain text conventions for certain formatting, which can be easily parsed later – but don’t have to be parsed to be understood by a regular human.
– TXT is really the only file format that meets every need I listed above, for a portable writer who wants to work anywhere, any time, and never have their text locked in to an irretrievable machine-readable format. Of course, TXT misses the key formatting assists which Markdown provides.
Markdown is a very human response, providing a minimally complex solution to an intractable problem with writing on electronic devices. Like any good idea, it can be made unwieldy by fracturing and feature creep. But the need is real. That’s why, over the past five years, there has been continual pressure from **actual users** for developers to actively embrace Markdown in their tools. Even though the whole point of Markdown is that “we can do it ourself with plain text.”
TorontoDave,
True. It is a universal format that can be edited by any tool, and infinitely extensible. And given it is plain text any dialects do not need to be interchangeable. Because even without tools humans can understand it.
So, you put a diagram that requires DOT, or Mermaid? Even if your MD editor does not have the capability to “render” it to a visual chart, it is extremely easy to read the structure as a human.
Markdown is very useful as a technical note sharing tool, and it will be here to stay for a long while.
(Note: the “marks” like *for italics* or 1. 2. 3. for lists have been used by humans long before markdown was a thing. Its syntax is natural)
edit: (sample but still not TT tags here!)
sukru,
Side debate:
You can call it italic, but when I read/write *text* in the context of human text it means bold. For example if I read *ahhh* in a screen play I would be thinking *ahhh* and not *ahhh*. This probably doesn’t matter much for most text, but since it carries semantic meaning in the context of a markup language then people may have different feelings about what it should map to.
Also, I don’t consciously know why, but for as long as I can remember this is how I’d naturally write a list in text…
1)
2)
3)
I hate markdown (and relatives) implicit usage with a passion: the main online git tools (github/gitlab) automatically format using markdown. I format plain text when writing pull request descriptions, markdown makes this hard.
I can no longer only think about the content, but also how the tool will interpret it, making me constantly switch between preview and the text itself. And don’t EVER interpret git commit message formatting.
Plain text is WYSIWYG.
asciidoc and rst are both objectively better. I hate that markdown has so much inertia.